Processes and Policy for Revising Monthly Production Statistics (GDP) at Statistics Canada

1
Processes and Policy for Revising Monthly
Production Statistics(GDP) at Statistics Canada

• Prepared by Michel Girard Erika Young
• Presented by Michel Girard, Statistics Canada
• OECD Short-Term Economic Statistics Working Party
• 25 27 June 2007
• Paris, France

2
Overview

• Revision policies
• Source data and monthly GDP
• Integrated revision process
• Frequency
• Reasons
• Computer system
• Clients

3
Revision policiesSource data

• Raw and seasonally adjusted data are normally
  revised over a 3-month period
• Respondent errors/availability of data
• Annual revision process for monthly data
• Coding on central registry
• Processing
• Accumulation of micro data corrections
• Data corrections in other programs
  (exports/imports, prices, administrative data)
• Annual surveys previous year is normally subject
  to revision
• Discrepancy between monthly and annual estimates
  becoming a concern
• Historical process for monthly data
• X12, NAICS, Profiling

4
Revision policiesSource data

• Issues
• Carrying back seasonal adjustments further back
  in time
• Micro vs macro adjustments
• Reconciling/benchmarking monthly to annual
  results
• Will require modifying processes and systems
• Calendarization
• Aligning source data and SNA revision policies

5
Revision policiesSource data

• Project on revision policy
• Description of policies by surveys
• Looking at revision policies in other NSO
• Extent to which policies fits the needs of
  clients including SNA
• Recommendations
• Notably reconciliation and benchmarking issues

6
Revisions to Monthly GDP

• According to frequency
• Monthly, quarterly, annual, historical
• Classification of revisions
• Source data
• Seasonality (including trading-day factors)
• Reconciliation
• Changes to methodologies
• Benchmarking including rebasing
• Classifications
• SNA conceptual revisions

7
Process of revision
Monthly Jan to Dec Quarterly Feb, May, Aug, Nov Annual September Historical 10 years
In August, revision back to January of previous year In September, revisions are carried back 5 years Tend to go back to 1961
Source data Seasonal Reconciling with quarterly GDP Incorporating quarterly source data Weights-prices Trading-day Methodologies Rebasing Benchmarking Central frame SNA 1993 Classifications
8
Computer System

• Menu driven, accessible by many users
• Charts/Reports
• Methodologies compiled for specific period of
  time
• Database can be decomposed into 4 categories
• Current version
• Revised version
• Published version
• Test/historical version
• Feeder system vs published results
• Avoiding building seasonality

9
Revisions and dissemination

• Qualitative information
• Source data
• Explained in the context of IO and IE Accounts
• Methodology
• Quantitative information
• Always growth rates and levels
• Sometime mean and standard deviations
• Occasionally bias, dispersion, concordance of
  movements

10
Clients

• Survey in 2004
• Satisfaction vis-à-vis aggregates 77
• Concern about detail 33
• /- 0.2 good trade-off on average for
  aggregates
• Accuracy more important than timeliness Data
  released 60 days after reference month
• Appreciate qualitative information, especially
  when special events take place
• Demand to incorporate revisions sooner than later
  for modelling and forecasting purposes
• Necessity to revise back when story do not change
  being questioned